Economics and Policies for Carbon Capture and Sequestration in the Western United States: A Marginal Cost Analysis of Potential Power Plant Deployment
by
Gary Shu
B.S., Electrical Engineering, Columbia University, 2003 M.A., Electrical Engineering, Princeton University, 2005
Submitted to the Engineering Systems Division and the Department of Urban Studies and Planning in Partial Fulfillment of the Requirements for the Degrees of
Master of Science in Technology and Policy and
Master in City Planning
at the Massachusetts Institute of Technology, February 2010 2010 Massachusetts Institute of Technology. All rights reserved. Signature of Author:
Technology and Policy Program, Engineering Systems Division Department of Urban Studies and Planning, January 15, 2010
Certified by:
Howard J. Herzog
Senior Research Engineer, MIT Energy Initiative Thesis Supervisor
Certified by:
Mort D. Webster Assistant Professor of Engineering Systems Thesis Supervisor
Certified by:
Karen R. Polenske
Professor of Regional Political Economy and Planning Thesis Supervisor
Accepted by:
Dava J. Newman
Professor of Aeronautics and Astronautics and Engineering Systems Director, Technology and Policy Program
Accepted by:
Joseph Ferreria, Jr.
Economics and Policies for Carbon Capture and Sequestration in the Western United States: A Marginal Cost Analysis of Potential Power Plant Deployment
by
Gary Shu
Submitted to the Engineering Systems Division and the Department of Urban Studies and Planning on January 15th, 2010 in Partial Fulfillment of the Requirements for the Degrees of
Master of Science in Technology and Policy and Master of City Planning ABSTRACT
Carbon capture and sequestration (CCS) is a technology that can significantly reduce power sector greenhouse gas (GHG) emissions from coal-fired power plants. CCS technology is currently in development and requires higher construction and operating costs than is currently competitive in the private market. A question that policymakers and investors have is whether a CCS plant will operate economically and be able to sell their power output once built. One way of measuring this utilization rate is to calculate capacity factors of possible CCS power plants. To investigate the economics of CCS generation, a marginal cost dispatch model was developed to simulate the power grid in the Western Interconnection. Hypothetical generic advanced coal power plants with CCS were inserted into the power grid and annual capacity factor values were calculated for a variety of scenarios, including a carbon emission pricing policy.
I demonstrate that CCS power plants, despite higher marginal costs due to the operating costs of the additional capture equipment, are competitive on a marginal cost basis with other generation on the power grid at modest carbon emissions prices. CCS power plants were able to achieve baseload level capacity factors with $10 to $30 per ton-CO2 prices. However, for investment in
CCS power plants to be economically competitive requires that the higher capital costs be recovered over the plant lifetime, which only occurs at much higher carbon prices. To cover the capital costs of first-of-the-kind CCS power plants in the Western Interconnection, carbon emissions prices have been calculated to be much higher, in the range of $130 to $145 per ton-CO2 for most sites in the initial scenario. Two sites require carbon prices of $65 per ton-CO2 or
less to cover capital costs. Capacity factors and the impact of carbon prices vary considerably by plant location because of differences in spare transmission capacity and local generation mix.
Thesis Supervisor: Howard J. Herzog
Senior Research Engineer, MIT Energy Initiative Thesis Supervisor: Mort D. Webster
Assistant Professor of Engineering Systems, Engineering Systems Division Thesis Supervisor: Karen R. Polenske
ACKNOWLEDGEMENTS
After two and half years of sometimes long, often hard work at MIT, I am grateful to a great many people for helping support me through times good and bad, and to finally getting me to where I want to go. (Wherever that may be.)
I am indebted to Howard Herzog and Mort Webster for the opportunity to work in this burgeoning field of energy. They had the confidence in my abilities and gave me the freedom to explore my research while directing my efforts for maximum impact. Karen Polenske gave me the belief that I have the ability and skills to make a contribution in the field of urban planning, a completely new area of expertise for me.
I am appreciative of the West Coast Regional Carbon Sequestration Partnership (WESTCARB), the MIT Carbon Sequestration Initiative, and the MIT Energy Initiative for their generous financial support of this work.
Thank you to Technology and Policy Program (TPP), the Engineering Systems Division (ESD), and the International Development Group (IDG) in the Department of Urban Studies and Planning (DUSP). I have had the opportunity to spend time in such a vibrant learning environment because these programs believed that I had something to offer to the MIT ecosystem and that I can ably become a future representative. I hope I have not let them down.
To the countless friends I have made who have helped keep me sane, motivated and hungry for more, I cannot thank you all enough. Needless to say, without your companionship I would not have had the fun and education that I have received here.
To my labmates, I need to thank you for your conversation and camaraderie in tackling all things energy: Ashleigh Hildebrand, Michael Hamilton, Holly Javedan, Yamama Raza, Manya Ranjan, Weifeng Li, Sarah Bashadi, Ellie Ereira, and Michael Follmann.
Special recognition is in order for Mary Gallagher, who has tirelessly supported us. She has made me feel welcome at MIT and in a research group, a considerable feat. It would not be the same without you.
To my family, I offer a few words. My brother, Anthony, I thank for the daily conversation, television access, and for living up to my parents’ expectations for me. I would like my grandmother to understand that I appreciate her care and worry.
TABLE OF CONTENTS
Acknowledgements... 4 Table of Contents... 5 List of Figures ... 7 List of Tables ... 9 List of Acronyms ... 10 1. Introduction... 121.1. Motivation: Climate Change... 12
1.2. Objectives and Scope... 14
2. Background ... 17
2.1. Prior Research and Studies ... 17
2.2. Carbon Capture and Sequestration... 20
2.3. Policies... 22
2.3.1. Regulation ... 22
2.3.2. Carbon Tax... 23
2.3.3. Cap-and-Trade Program... 24
2.3.4. Subsidies and Demonstration Projects... 26
2.4. Electricity Grid... 27
2.5. Electricity Markets... 28
2.5.1. Regulation ... 28
2.5.2. Restructuring and Competitive Markets ... 29
2.5.3. Electricity Costs ... 31
2.6. Area of Interest – Western Interconnection ... 31
3. Methodology ... 34
3.1. Dispatch Model... 34
3.2. Dispatch and Capacity Factor Calculation... 38
3.3. Marginal Costs ... 40
3.3.1. Data Sources ... 40
3.3.2. Calculation ... 40
3.4. CCS Power Plants ... 42
3.5. Discussion of Hypothetical Power Plant Sites... 43
3.5.1. Central California... 45
3.5.2. Bordering California ... 46
3.5.3. Northern California... 47
3.5.4. Northwest... 48
3.5.5. Four Corners ... 49
3.6. Simulation Scenarios Summary... 51
4. Results and Analysis ... 52
4.1. Dispatch Model Simulation Results... 53
4.1.1. Carbon Price Scenarios ... 53
4.1.2. Carbon Prices with CCS Power Plants Scenarios... 55
4.1.3. Alternative Fuel Price Scenarios... 58
4.1.6. Partial Capture CCS Power Plants Scenario... 72
4.2. Financial Analysis of the Investment Decision ... 75
4.2.1. Base Case Financial Analysis ... 77
4.2.2. High Natural Gas Price Financial Analysis ... 80
5. Discussion and Policy Implications ... 82
5.1. Carbon Emissions Price ... 82
5.2. Siting and Transmission... 84
5.3. Technology Support Policies ... 85
6. Conclusions and Future Work ... 87
6.1. Transmission Constraints and the Regional Fuel Mix Matter ... 87
6.2. CCS Power Plants Will Dispatch With Modest Carbon Prices ... 87
6.3. Investment for First-of-a-Kind CCS Generation Require High Carbon Prices ... 88
6.4. CCS Technology Deployment Policy and Siting Strategies: Policies to Recover Capital Costs and Coal Regions ... 88
6.5. Future Work ... 89
References... 91
Appendices... 95
Appendix A: Dispatch and Capacity Factor Calculation Example ... 95
Appendix B: Marginal Cost Calculation Example ... 96
Appendix C: Base Case Scenario Inputs ... 97
Appendix D: Alternative Fuel Price Scenario ... 98
Appendix E: CCS Power Plant Efficiency Scenario ... 100
Appendix F: CCS Capture Rate Sensitivity Scenario... 102
LIST OF FIGURES
Figure 2-1 Status of Electricity Restructuring by State, September 2009 ... 30
Figure 2-2 Electrical Interconnections in the United States and Canada... 32
Figure 3-1 WECC Power Plants Represented in the Dispatch Model... 36
Figure 3-2 Load Duration Curve for 2005 CAISO System Area ... 39
Figure 3-3 Hypothetical CCS Power Plants Modeled in the Western Interconnection. ... 43
Figure 4-1 (a) Capacity Factor Increase Weighted by Maximum Capacity due to a $100 per ton-CO2 Carbon Policy Relative to No Policy ... 53
Figure 4-2 (b) Capacity Factor Decrease Weighted by Maximum Capacity due to a $100 per ton-CO2 Carbon Policy Relative to No Policy ... 54
Figure 4-3 Capacity Factor vs Carbon Price for Individual Hypothetical CCS Power Plants ... 56
Figure 4-4 Capacity Factor for Individual Hypothetical CCS Power Plants. ... 57
Figure 4-5 Capacity Factor vs Carbon Price for CCS Power Plant at Midway Site for Fuel Price Scenarios ... 59
Figure 4-6 Capacity Factor vs Carbon Price for CCS Power Plant at Four Corners Site for Fuel Price Scenarios... 60
Figure 4-7 Capacity Factor vs Carbon Price for CCS Power Plant at Inyo Site for Fuel Price Scenarios ... 60
Figure 4-8 Capacity Factor for Individual Hypothetical CCS Power Plants in High Coal Price, Low Natural Gas Price Scenario... 62
Figure 4-9 Capacity Factor vs Carbon Price for CCS Power Plant at Four Corners Site for Efficiency Sensitivity Scenarios ... 64
Figure 4-10 Capacity Factor vs Carbon Price for CCS Power Plant at Vincent Site for Efficiency Sensitivity Scenarios... 65
Figure 4-11 Capacity Factor for Individual CCS Power Plants with 17,000 BTU per kWh Heat Rate. ... 66
Figure 4-12 Capacity Factor vs Carbon Price for CCS Power Plant at Midway Site for Capture Rate Sensitivity Scenarios... 67
Figure 4-13 Capacity Factor vs Carbon Price for CCS Power Plant at Cholla Site for Capture Rate Sensitivity Scenarios... 68
Figure 4-14 Capacity Factor for Individual CCS Power Plants with 50% Capture Rate. ... 69
Figure 4-15 Capacity Factor vs Carbon Price for CCS Power Plant at Burns Site for Capture Rate Sensitivity Scenarios... 70
Figure 4-16 Capacity Factor vs Carbon Price for CCS Power Plant at Midway Site for Partial Capture Scenarios ... 73
Figure 4-17 Capacity Factor vs Carbon Price for CCS Power Plant at Cholla Site for Partial Capture Scenarios ... 74
Figure 4-18 Capacity Factor for Individual CCS Power Plants with 0% Capture and 8,500 Heat Rate. ... 75
Figure 4-19 Capacity Factor and Income vs. Carbon Price at a Four Corners CCS Power Plant 78 Figure 4-20 Capacity Factor and Income vs. Carbon Price at a Midway CCS Power Plant ... 79
Figure A-1 Capacity Factor vs Carbon Price for Individual Hypothetical CCS Power Plants – High Coal-High Natural Gas Fuel Price Scenario ... 98
Figure A-2 Capacity Factor vs Carbon Price for Individual Hypothetical IGGC Power Plants – Low Coal-High Natural Gas Fuel Price Scenario... 99 Figure A-3 Capacity Factor vs Carbon Price for Individual Hypothetical IGGC Power Plants –
High Coal-Low Natural Gas Fuel Price Scenario... 99 Figure A-4 Capacity Factor vs Carbon Price for Individual CCS Power Plants with 13,500 BTU
per kWh Heat Rate... 100 Figure A-5 Capacity Factor vs Carbon Price for Individual CCS Power Plants with 17,000 BTU
per kWh Heat Rate... 101 Figure A-6 Capacity Factor vs Carbon Price for Individual CCS Power Plants with 50% Capture
Rate. ... 102 Figure A-7 Capacity Factor vs Carbon Price for Individual CCS Power Plants with 0% Capture
Rate. ... 102 Figure A-8 Capacity Factor vs Carbon Price for Individual CCS Power Plants with 0% Capture
Rate and 8,500 BTU per kWh Heat Rate... 103 Figure A-9 Capacity Factor vs Carbon Price for Individual CCS Power Plants with 50% Capture
Rate and 10,000 BTU per kWh Heat Rate... 104 Figure A-10 Capacity Factor vs Carbon Price for Individual CCS Power Plants with 60%
Capture Rate and 8,500 BTU per kWh Heat Rate. ... 104 Figure A-11 Capacity Factor vs Carbon Price for Individual CCS Power Plants with 70%
Capture Rate and 10,550 BTU per kWh Heat Rate. ... 105 Figure A-12 Capacity Factor vs Carbon Price for Individual CCS Power Plants with 80%
Capture Rate and 10,850 BTU per kWh Heat Rate. ... 105 Figure A-13 Capacity Factor vs Carbon Price for Individual CCS Power Plants with 90%
LIST OF TABLES
Table 3-1 Hypothetical CCS Power Plant Site Summary... 44
Table 3-2 Summary of Dispatch Model Scenarios ... 51
Table 4-1 Key Financial Assumptions Applied in Capital Cost Evaluation ... 76
Table 4-2 Carbon Prices for Profitable CCS Power Plant Investment at Each Site: Base Case... 77
Table 4-3 Carbon Prices for Profitable CCS Power Plant Investment at Each Site: Base Case and High Natural Gas Price Scenarios... 80
Table A-1 Capacity Factor Calculation Example for Hypothetical CCS Power Plant from Sampled Dispatch ... 95
Table A-2 Example of Marginal Cost Calculation with Sources ... 96
Table A-3 Base Case Fuel and Generation Inputs ... 97
Table A-4 Base Case CCS Power Plant Inputs... 97
Table A-5 Alternate Fuel Price Scenario Inputs ... 98
Table A-6 CCS Power Plant Efficiency Scenario Inputs... 100
LIST OF ACRONYMS
AC Alternating Current
ACES American Clean Energy and Security Act
BPA Bonneville Power Authority
BTU British thermal units
CAIR Clean Air Interstate Rule
CAISO California Independent System Operator
CCS Carbon capture and sequestration
CGE Computable general equilibrium
CO2 Carbon dioxide
DC Direct Current
DOE United States Department of Energy
ED Economic dispatch
eGRID Emissions and generation resource integrated database
EIA United States Energy Information Administration
EOR Enhanced oil recovery
EPA United States Environmental Protection Agency EPPA Emissions Prediction and Policy Analysis model
ETS European Union Emission Trading Scheme
FERC United States Federal Energy Regulatory Commission
FOAK First-of-a-kind
GHG Greenhouse gas
HHV High heating value
IEA International Energy Agency
IGCC Integrated gasification combined cycle
IPM Integrated Planning Model
ISO Independent system operator
kV Kilovolt
kW Kilowatt
kWh Kilowatt-hour
LMP Locational marginal price
MARKAL Market Allocation model
MIT Massachusetts Institute of Technology
MMBTU Million British thermal units
MW Megawatt
MWe Megawatt-electrical
MWh Megawatt-hour
NEMS National Energy Modeling System
NOAK Nth-of-a-kind
NOx Nitrogen oxide
OPF Optimized power flow
PC Pulverized coal
PRB Powder River Basin
PTC Production tax credit
PUC Public utility commission
RGGI Regional Greenhouse Gas Initiative
RTO Regional Transmission Operator
SCOPF Security constrained optimized power flow
SCPC Supercritical pulverized coal
SOx Sulfur oxide
Ton-CO2 Short ton of carbon dioxide
WECC Western Electricity Coordinating Council
1. INTRODUCTION
1.1. Motivation: Climate Change
Climate change from carbon dioxide emissions and other greenhouse gases is altering the planet's environment and atmosphere. Possible direct climate impacts include extreme weather changes, rising sea levels and the increase in catastrophic weather events. The indirect impacts include drought and flooding on land, severe alterations in local ecosystems and the destruction of infrastructure.
Potential reductions in the likelihood of catastrophic environmental damage from anthropogenic climate change requires a decrease in carbon dioxide (CO2) emission to the atmosphere and the
stabilization of CO2 and other greenhouse gases (GHG) concentrations. This would entail a
dramatic reduction in CO2 emissions from current levels. There are numerous proposals for
large reductions in emissions up to 80% below current levels by midcentury. (EIA 2008; Stern 2006; Alley et al. 2007)
GHG emissions are produced in many different sectors of the human economy and society; however, the energy sector in particular provides a large contribution to greenhouse gas emissions and to possible climate change. The combustion of fossil fuels like coal and natural gas in electrical generators are a significant source of carbon dioxide emitted into the atmosphere as a result of human activity. Mitigation of climate change will necessarily include reduction in CO2 from the electricity sector.
In the United States, the power sector is responsible for over 40% of all energy-related CO2
emissions, with emissions from coal-fired power plants accounting for 80% of electricity-related emissions. Using coal for electricity thus accounts for nearly a third of the United States’ CO2
emission alone. (EIA 2008) If United States, along with other countries, is to decrease its emissions by 80%, carbon emissions from coal-fired power plants must be reduced.
Carbon capture and sequestration (CCS) has been identified as a promising set of technologies with the potential for reducing carbon emissions from coal-fired power plants. (IPCC Special
Report on CCS 2005) CCS captures CO2 from a power plant’s flue gas and pumps concentrated
CO2 into geological storage reservoirs, preventing its release into the atmosphere. (IPCC Special
Report on CCS 2005; MITEI 2007) The widespread adoption and deployment of CCS would allow the continued use of coal, an abundant and inexpensive fossil fuel resource, with a much smaller carbon footprint.
Electricity generation with CCS would have much higher costs than current generation options and is not sufficiently mature to compete with other types of electrical generation. (Hamilton et al. 2008) CCS also requires sequestration sites in which carbon dioxide would be stored, necessitating additional infrastructure costs and siting constraints. Coal-fired power plants with CCS require large capital expenditures and initial outlays making it difficult for rapid and widescale deployment.
In the future, regulation may be imposed on the emission of carbon dioxide and other greenhouse gases. A price on carbon, either through a tax or through the market of a cap-and-trade permit market, will impact the cost of electricity generated by fossil fuels proportionate to their CO2
emissions rates. Given that coal emits approximately twice as much carbon dioxide per kilowatt-hour as natural gas does, a government policy to price greenhouse gas emissions will favor natural gas power plants and coal-fired power plants with greenhouse gas control technology like CCS. (Freese & Clemmer 2006)
The determination of which power plants operates is made through a dispatch calculation by the transmission operator. Regions as small as a single utility’s balancing area or as big as the operating area of a Regional Transmission Operator (RTO) use dispatch. In a dispatch
calculation, power plants in an area are turned on and off in order to minimize the cost of serving electricity demand within operational constraints like transmission congestion. With little to no transmission congestion, the cheapest power plants get turned on the most.
The imposition of a carbon price will cause a dispatch calculation to favor less carbon-intensive fuel sources. Because CCS power plants are more expensive to operate, CCS power plants would be disadvantaged in a dispatch calculation in the absence of supporting policies like a
carbon emissions price. Even if policies exist that favor a CCS power plant to dispatch, transmission constraints can make it physically infeasible to generate electricity from a given source. These constraints vary by location and, for a given location, depend on the state of the entire interconnected system which is constantly changing at each point in time.
Over an entire year, the percentage of time that a generator will operate, its “capacity factor”, is a critically important input to financial calculations that investors use to determine the profitability of a given project. A generator must profitably run for enough hours of the year to recover both the initial capital costs of a construction and its marginal operating costs while still providing a profit to its owner.
Most studies of CCS technology and specific projects assume a set capacity factor. However, given the nonlinear physical constraints of the electric power system, it is important to determine whether a CCS power plant would dispatch even if it is cheaper than other generation in pure economic terms, and to explore how constraints like transmission capacity depend on the specific siting within the power system.
In order to deploy CCS technology successfully, it is important to understand how it operates and dispatches within the electric transmission grid.
1.2. Objectives and Scope
The purpose of this study is to understand the operation and utilization of CCS power plants in a realistic electric transmission grid. This includes how CCS power plants respond to varying operating conditions like lower efficiencies and capture rates, varying market conditions like different fuel price scenarios, and varying government policies regarding carbon emissions pricing. I will model CCS power plants in different locations to investigate how siting affects operation through transmission congestion and regional competition.
Our analysis will perform a transmission-constrained generation dispatch simulation using an optimization method employed by transmission organizations called optimized power flow (OPF) to calculate the economics and utilization of potential CCS power plants. By exploring
how the transmission network affects the economics of CCS power plants under various scenarios, I consider a deployment strategy that prioritizes effective policies and locations for power plants with CCS technology can be considered.
The area of study is the Western Interconnection of North America, encompassing a vast region of the Western United States, two Canadian provinces, and a small section of Baja Mexico. This region, overseen by the Western Electricity Coordinating Council (WECC) represents an
extremely broad area with diverse regional energy mixes. Transmission is a concern in this region because the major load centers in California import substantial amounts of electricity. The WECC is also an area where active CCS projects are currently planned or underway, and the study of siting strategies is important and can have an impact.
I will examine alternative scenarios will be explored to determine the conditions, political and economic, that would support the development of CCS projects. Scenarios include market fuel prices since these affect the costs of electricity produced by other power plants on the grid that would compete with any CCS plant. By treating natural gas and coal as substitutable fuels in the production of electricity, shifting the prices of these fuels will lead to corresponding adjustments in the generation mix used to meet electricity demand. Because of the lower CO2 emissions rates
from gas-fired generation, for CCS to be economic for wide-scale commercial use, the electricity generated by a coal-fired power plant with CCS will need to be cost-competitive with the
electricity generated from natural gas.
Transmission limitations are also crucial in the siting and locations of power plants. As electricity is supplied to the transmission grid, transmission constraints can limit the ability to transfer power from generation to load. The demand for electricity is generally concentrated in large metropolitan urban areas like Las Vegas, Phoenix, and Southern California in the west. However, land-use zoning, permitting and real-estate costs in these dense urban areas generally require power plants to be built far away from population centers, especially larger power plants that produce electricity at lower cost. Utilities may need to build large, expensive transmission lines may need to be built to allow electricity from these power plants to be delivered to urban load. In the absence of new transmission, plants at some locations may not dispatch even if they
are competitive on a purely marginal cost basis. In this analysis, I explore the sensitivity of results to location by testing 14 different sites.
Calculated capacity factors for different locations and conditions will then be used as an input to a financial investment decision analysis, determining the conditions (e.g. carbon prices, fuel prices, etc.) and the locations where it is economical to build a CCS power plant. The results provide insight into policies that are likely to encourage such investments.
The remaining chapters are organized as follows. Introduction and background into CCS technology, policy tools, and the electric grid are provided in Chapter 2. Chapter 3 offers detailed discussion into assumptions, data, methodology and construction of the marginal cost dispatch model. In Chapter 3, specific attributes and history of the CCS sites chosen to be modeled are described. Chapter 4 contains a comprehensive analysis of the simulation results, and individual scenarios and their consequences are explained. A discussion of the implications for policy and business strategy is provided in Chapter 5. Finally, Chapter 6 presents my
2. BACKGROUND
2.1. Prior Research and Studies
The energy economics literature contains numerous studies and analyses of advancing the deployment of carbon capture and sequestration in the electricity sector. These studies often offer perspectives from an economic or technological analysis in order to inform policymakers, businesses, governments, and organizations on potential pathways or barriers to CCS adoption. A brief review of major studies and methodologies follows in order to provide context and motivation for the novel approach and analytical results regarding plant capacity factors and investment cash flow decisions this thesis offers.
The highest abstracted level of studies regarding CCS analyzes the technology’s attributes to compare CCS against alternatives in the economic, political and policy environments in order to produce next-step recommendations to advance the state of the art. For instance, the World Resource Institute (WRI) offers recommendations for disparate areas like risk assessment, site selection, injection operation. (WRI 2008) The International Energy Agency (IEA) gives suggestions on techniques to incentivize CCS deployment in the international sphere. (Philibert et al. 2007) The MIT Future of Coal Report recommends a special emphasis on ensuring that CCS demonstration projects are rapidly developed worldwide. (MITEI 2007).
These policy studies offer actionable recommendations using aggregated information in order to identify legal, economic, and policy barriers that may need attention. However, analyses of novel technologies like CCS require projections of future use. Such projections are made using models that can take a variety of forms, including economy-wide, energy sector, electric capacity planning, unit commitment, and generator dispatch model. Each type of model is designed and used to answer specific questions that can provide insight from a different perspective into how best to deploy a technology like CCS.
One of the most important questions that economic models answer is what the impact of a technology or a technology portfolio will be on supply, demand and prices. In order to answer such questions, partial equilibrium or computable general equilibrium (CGE) models are applied
to simulate a sector or region's economy. One example is the US DOE EIA National Energy Modeling System (NEMS) used in the EIA’s Annual Energy Outlook, which splits the United States into nine Census regions which, due to variations in weather and fuel supply accessibility, contain their own characteristic demand and prices for certain types of energy as well as separate electricity markets. (US Energy Information Administration 2009) The International Energy Agency (IEA) uses a similar structure for the modeling behind its influential annual World Energy Outlook report. (International Energy Agency 2009) The MIT Emissions Prediction and Policy Analysis (EPPA) model is a CGE model of the global economy, consisting of 16
geopolitical regions with trade among them, and each made up of economic sectors with greatest resolution in energy fuels, technology, and consumption. (Sergey Paltsev et al. 2005)
Economic models can identify how large, economy-wide impacts; for example, how the
additional coal input into CCS power plants will affect coal and related markets and not just the impact of coal-fired power plants with CCS in the electric power sector. However, since these kinds of top-down models must necessarily account for large sectors of economic activity, analysts often omit details regarding certain sectors and technologies - like the power sector and coal thermal generation with CCS - or aggregate them into phenomenological variables due to complexity and computational concerns. One particularly important technological attribute I attempt to determine is the nonlinearity of the electricity transmission system and how it affects the usage rate, the "capacity factor", of a coal-fired power plant with CCS.
One alternative is a more detailed technological model. These models include greater resolution of technological options, each with many associated technology-specific attributes in order to compare technologies against one another. The usage and potential revenues of an electricity power plant are determined by the technologies that are chosen by the model. MARKAL (MARKet ALlocation) is well-known example of a technology optimization model. The advantage of such an approach is that the model can choose from a comprehensive suite of technologies according to cost and performance characteristics. However, a major disadvantage in these models is the lack of feedbacks, like changes in market prices with large technology usage shifts.
A significant attribute of the power sector that many of these types of models often lack is detail regarding the transmission system. Transmission constraints caused by congestion can limit the ability for even the cheapest and environmentally-friendly electric generation technology to deliver energy, a crucial detail. More aggregated models often lack even a rudimentary transmission system and often require exogenous penetration rates that limit the level of technology adoption to the realm of feasibility.
One example of a model that incorporates transmission detail is one that the US EPA and the FERC sometimes use from an external contractor, ICF International, who provides technical modeling expertise using their own economic model, the Integrated Planning Model (IPM). This model contains detailed network information on the national transmission grid as well as the location, type, and costs of individual power plants.
Utilities and transmission agencies often use more detailed transmission and generation computer models to run the power grid. Capacity expansion planning models optimize long-term additions to the power grid using cost and technological attributes of possible new generation using cost and reliability as primary factors. (Hobbs 1995; Kagiannas et al. 2004) Unit commitment models look at how generation utilization must be planned ahead of time in order to optimize scheduled maintenance outages and to work around technological attributes such as ramp rates and start-up costs for thermal power plants. (Sen & Kothari 1998) These more detailed models still do not explicitly take into account the cost of electricity at a particular location nor do these models identify constraints in the ability to deliver energy over
transmission lines. Both are crucial factors in determining the utilization and capacity factor rates in an electric power generator.
An optimal power flow (OPF) dispatch model can explicitly calculate the cost-effectiveness of a power plant while considering transmission and other physical constraints and the existing cost of electric power on the grid. For new technologies such as CCS, understanding the conditions in which a potential generator is cost-competitive is crucial to developing that technology at minimal cost. Because it also requires siting power plants in particular locations on the power grid network, OPF models can also reveal location-specific effects of generation. CCS, with its
need to sequester CO2 in specific geologic formations, is also very location-dependant and a
dispatch model can explore the interactions between transmission and sequestration siting.
The OPF dispatch model has previously been applied to examine various aspects of integrating advanced coal-fired power plants into electricity systems. Studies have represented dispatch through simplified assumptions in plant-level cost analysis as well as CGE models. (Rubin et al. 2007; McFarland & Herzog 2006; Geisbrecht & Dipietro 2009) Others have used a dispatch calculation to look at how coal-fired power plants with CCS would fare under conditions such as GHG caps and other policies. (Johnson & Keith 2004; Wise & Dooley 2009) Studies have also specifically examined the dispatch characteristics for technology penetration in the Western Interconnection. (Ford 2008)
I will expand on previous studies in this thesis by using a marginal cost dispatch model to capture detail regarding the existing transmission grid and specific locations in the Western United States for advanced coal power plants. Using this model, I will explicitly calculate the availability and cost-effectiveness of a hypothetical coal-fired power plant with CCS. In this way, important transmission constraints will be captured that can determine much of the usage and revenue of a power generator which can cause significant variations across transmission grid locations. The impact of policies, future scenarios, and technology characteristics will thus incorporate a spatial component according to a hypothetical plant's location in the simulated transmission grid.
2.2. Carbon Capture and Sequestration
While CCS technology can be used in natural gas power plants, refineries, cement manufacturing and chemical processing plants, the application of CCS in coal-fired power plants represents some of the greatest carbon mitigation potential. The current state of CCS in coal-fired power plants uses two predominant designs: pre-combustion and post-combustion capture.
The technology that is most compatible with the existing fleet of coal-fired power plants (i.e., pulverized coal (PC)) is post-combustion CCS which removes CO2 from a power plant’s flue
gas. Post-combustion systems afford the ability to fit CCS systems onto existing coal-fired power plants, allowing retrofits into the existing fleet but typically at a heavy energy penalty.
Pre-combustion systems can be used where coal is first gasified, like Integrated Coal
Gasification Combined Cycle (IGCC) power plants. Since the CO2 in gasified coal is under
pressure and not diluted by combustion air, the cost of capturing CO2 can be substantially less.
However, IGCC power plants have not yet proven to be commercially viable on a large-scale. IGCC technology is also highly integrated; a failure in a component of the system will likely lead to the unavailability of the entire plant.1
The cost of CCS technology when compared to reference plants can be substantial. Nth-of-a-kind (NOAK) supercritical pulverized coal (SCPC) power plants have been found to have costs of over $3,000 per kWe, representing an additional cost of about 50% over plants without CCS equipment. (Hamilton et al. 2008) Studies into the implementation of post-combustion retrofit CCS system on existing PC power plants has found that the capital cost by itself would be half a billion dollars. (MITEI 2009, p.21) The MIT Energy Initiative retrofit workshop report found that for Nth-of-a-kind (NOAK) PC power plants with CCS, CO2 avoidance costs would be in the
range of $50 to $ 90 per ton-CO2. (MITEI 2009) Costs for first-of-a-kind (FOAK) PC CCS
power plants would be greater than $100 per ton-CO2. (Al-Juaied & Whitmore 2009)
Costs for IGCC and oxy-fuel combustion are highly tentative and subject to large amounts of uncertainty. Nevertheless, other recent studies have calculated that, in the case of a first-of-a-kind (FOAK) IGCC power plant, capital costs are in the range of $6,000 to $7,000 per kWe. (Al-Juaied & Whitmore 2009) The reported costs of DOE’s FutureGen experimental CCS power plant fall within this range at $6,545 per kWe.2 (GAO 2009) While CCS has been determined to be a necessary technology to reducing overall CO2 emissions, its high costs
necessitate careful planning and understanding of its effects in order to maximize its impact with minimal cost.
1
In order to avoid releasing the CO2 into the atmosphere, CO2 must be stored. The most
promising method of CO2 storage is to pump pressurized CO2 into geological sinks. This is
currently being done for the benefit of enhanced oil recovery (EOR) in declining oil fields for financial benefit. However, the use of CO2 for EOR is limited to areas accessible to oil fields.
More common are saline aquifers that can be found in many regions of the United States and other countries.
Other costs involved in CCS include transportation of CO2 and the potential infrastructure cost of
building CO2 pipelines to connect CCS power plants with CO2 sinks. In order to minimize such
transportation infrastructure for FOAK CCS power plants, siting of power plants will probably be as close as possible to storage locations.
2.3. Policies
Governments have at their command a set of policy instruments to diminish CO2 emissions from
the power sector in general and in CCS in particular. Each has their advantages and
disadvantages and particular implementation issues that greatly impact their effectiveness as a policy tool.3
2.3.1. Regulation
One method of ensuring that CCS power plants are used for coal-fired power plants is to simply mandate that all newly built coal-fired power plants include the technology. Such moratoriums on traditional coal-fired power plants have advanced in fits and starts but have not gained widespread traction. Advocates for such a moratorium famously include former Vice President Al Gore and former NASA atmospheric scientist James Hansen, who has claimed that adding even one additional coal-fired power plant to the world’s national grid can be the tipping point in the climate system. (Hansen 2008) To date, no national government has issued such a
moratorium on traditional coal-fired power plants and required the use of full capture CCS.
A method of regulating emissions is to set emissions performance standards on a per energy unit basis. California passed legislation (SB 1368) in 2007 setting a maximum emissions limit of
3
1100 lbs-CO2/MWh for any new power plant that provides power to the state. This emissions
rate is similar to that of lower-efficiency natural gas-fired power plants, about half of a typical coal-fired power plant, thus precluding the construction of future coal generation. The state of Washington passed similar legislation with SSB 6001.
Such regulations can be effective in ensuring that high CO2 emissions power plants will not be
built. However, they do not offer much financial incentive or support in building low-carbon sources nor do they ensure that new technology will be developed or used in future instances. One critique of such a policy tool is that power developers will merely shift heavily toward natural gas-fired power plants. They also deny the use of coal as a fuel in new conventional power plants no matter how cheap or convenient it may be. Other policy instruments may be more flexible in how they incentivize new technologies.
2.3.2. Carbon Tax
A carbon tax puts a cost on every ton of CO2 emitted. By placing a value on CO2 emission,
sources with higher emissions will have a higher cost than lower emissions sources. This will allow initially expensive low-carbon electricity generation to be more competitive against cheaper carbon-intensive sources of electricity like coal-fired power plants. Utilities and
developers will be incentivized to use to power plants that emit less CO2 like renewables, nuclear
and – relative to coal – natural gas.
How a carbon tax impacts the power sector can vary depending on how the tax is imposed but theoretically its effect should be the same. If the burden of paying the tax is imposed upstream in the energy process, such as the coal minemouth or the gas well, the cost of the carbon
emissions is encapsulated in the price that power plants pay for fuel. If a carbon tax is imposed instead on the power plants generating electricity and emitting the CO2 then the utility will have
to pay directly for its emissions. Although the overhead for administering the tax may be very different because of the large difference in sources covered, in both cases, costs will be passed through to electricity consumers and electricity prices will be proportionately impacted by the additional cost of CO2, in theory.
The advantage of a carbon tax lies in its transparency, certainty and ease of government
monitoring and enforcement. A carbon tax allows businesses and the markets to respond to the price on emissions and reallocate resources to minimize CO2 emissions as cheaply as possible.
Unlike other market-based regulations that rely on trading and exchanges, a carbon tax will provide a firm price for carbon emissions that is less alterable. A carbon tax also requires less overhead to administer for the government. (Babiker et al. 2003) Governments do not have to ensure that required equipment such as emission control technology are installed or manage permits. The supervising agency simply counts emissions and issues a bill.
The primary disadvantage of a carbon tax in the United States is a political one. According to one poll by GlobeScan, a majority of those polled in the United States are against carbon taxes. (Globescan 2007) The political realities make it difficult for politicians and legislators to back a carbon tax without significant public backlash. Carbon taxes as such are a non-starter in the United States as a carbon emissions regulation mechanism.
2.3.3. Cap-and-Trade Program
Cap-and-trade schemes have been proposed based on the experience of their application in SOx
and NOx emissions markets. In a cap-and-trade program, the total amount of some emissions is
set at a maximum level and allowances to emit are administered through government-monitored permits. The value of emissions permits can be determined by requiring some or all emissions permits to be distributed through a government-run auction and allowing the remained to be bought and sold on an exchange (the “trade” part of cap-and-trade), allowing a market to set the price of emissions.
CO2 cap-and-trade programs have been used to limit carbon emissions. The European Union
Emissions Trading Scheme (ETS) has been in operation since 2005, the largest such cap-and-trade program in the world, covering power and industrial sectors. The price that the ETS has put on a metric ton of carbon emission has varied considerably from €30 at its peak to a low of fractions of €1.4 In the United States, the Regional Greenhouse Gas Initiative (RGGI) recently started in 2009 in the Northeast between ten states to limit the emissions of the power sector.
4
The value of RGGI’s permits have varied from a high of $3.51 per (short) ton of CO2 to a low of
$1.87 per ton of CO2..
Major legislation is currently under consideration by the United States Congress to establish a nationwide cap-and-trade program. In Congress, H.R. 2545, the American Clean Energy and Security Act of 2009 (ACES, also nicknamed Waxman-Markey) has passed the House of Representatives and would establish a cap-and-trade program for CO2. At the beginning of
2010, S. 1733 the Clean Energy Jobs and American Power Act is being considered in the Senate. The major component of both of these programs is an economy-wide cap-and-trade program, including the electric power sector.
According to economic theory, a cap-and-trade program would have impacts similar to a carbon tax. For cap-and-trade, a CO2 emissions target is set and the market determines the carbon price.
For a carbon tax, a carbon price is administratively set and the market determines the level of CO2 emissions. Another difference is that in a cap-and-trade program, policymakers decide how
to allocate emissions permits while in a carbon tax policymakers decide how to allocate revenue from the emissions tax.
The ability to distribute emissions allowances through a political process can provide significant cost relief to emissions sources. In this way, the implementation of a cap-and-trade program – a major new regulation that would affect entire economies – can be made politically palatable. In the United States, for instance, the allocation of emissions permits to coal interests has allowed moderates from coal-dependant regions to be more inclined to back the cap-and-trade program. (Lerer 2009)
One of the major disadvantages of a cap-and-trade system is that a lax cap can create leakages in emissions and price volatility. For instance, the ETS program provided too many permits in its first phase which, when discovered, caused the value of permits to bottom out. RGGI has experienced similar problems as its auction results have produced lower prices. Volatility in the value of emissions allowances creates an uncertain investment environment for the deployment of new technologies. One remedy for this is a “price collar” that sets a floor and ceiling on the
value of emissions permits, a solution currently being discussed in conjunction with the Kerry-Boxer Senate bill. (Energy Washington Week 2009)
2.3.4. Subsidies and Demonstration Projects
Governments can use public funds to provide subsidies or to support demonstration projects for large unproven technologies like CCS or new nuclear power plants. Programs like the DOE’s Loan Guarantee Program provide large, low-cost loans to developers of power projects. Grants are also provided through offices like the Advanced Research Projects Agency – Energy (ARPA-E) to perform research and development and to support the commercialization of energy
technologies.
The initial risk of multi-billion dollar projects for early technology demonstrations could warrant such use of public money. The rationale is that private businesses and banks may be unwilling to support immature technologies with highly risky financial returns. But a larger public benefit may exist in advancing the state of the technology in cost and knowledge to the point that the private sector can use and deploy the technology. To that end, governments may take on risk through financing, insurance or the backing of debt.
Governments and industrial groups have been putting together FOAK demonstration plants for IGCC with CCS. FutureGen is a consortium of government and businesses to create the first commercial scale IGCC power plant with sequestration in the United States. The project has had a troubled history having hit roadblocks and funding droughts along with some project
restructuring. The United States government along with American Electric Power is also supporting the Mountaineer experimental project in West Virginia. In China, GreenGen is a project that started construction in the summer of 2009 and is expected to be operational
sometime during 2010 or 2011. Up and running in Germany is Schwarze Pumpe, a experimental project by utility company Vattenfall.
Subsidies can accelerate technology development and deployment by getting FOAK plants built quicker. The rationale for using public funding for technology development are the widespread benefits that private developers may have difficulty capturing. For instance, an industry as a
whole could benefit from the technical and operating learning furnished by the first set of technology demonstrations. Businesses, like the finance and insurance industries, may become more comfortable with supporting large, yet-unproven technology investments. Finally, the public will gain some familiarity and understanding of the benefits that a technology like CCS may have, and lend community and political support.
The primary disadvantage of subsidies is that they are often directed at specific technologies where success is not guaranteed. In these cases, not only are the subsidy funds lost, but there is also the opportunity cost of not using that money to support other technologies and other public investments. Subsidies can be viewed as calculated risks to support specific technologies in which the payoff of further technology development is uncertain.
2.4. Electricity Grid
The electricity grid is a large network of transmission lines connecting disparate power generation stations with load areas and the distribution network. In order for CCS or any new electricity technology to operate economically and profitably, it must compete against other generation types on the power grid: plants using natural gas, existing coal, diesel, geothermal, wind, solar and hydropower. Fossil fuel burning power plants also have a set of differing technologies that produce varying operating characteristics even though the plants are burning the same fuel.
Different regions will utilize different portfolios of technologies and fuels to produce electricity according to their resource endowment and proximity to fuel supplies. Other factors that may affect capacity expansion decisions include regional weather variations, transmission capacity, level of load, expected growth and environmental regulations, to name a few.
Capacity planners are also cognizant of the diurnal and seasonal cycles that require generation that can serve constantly varying demand. Power generation is categorized as baseload, intermediate, and peaking depending on which part of the load curve it serves. This categorization is determined by ramp rates and ability to dispatch but mostly by cost of generation. Baseload generation runs nearly all of the time with greater than 80% capacity
factors and peaking power plants can run as little as a few percent of the hours of the year. Intermediate or shoulder power plants run between peaking and baseload.
In order to serve growing electricity demand in the United States and the world while at the same time minimizing GHG emissions, there is a significant need for low-carbon baseload generation, like hydropower, nuclear and coal-fired power with CCS. Renewable sources, like solar and wind, suffer from intermittent generation and the inability to store and dispatch energy to meet periods of peak load.
2.5. Electricity Markets 2.5.1. Regulation
Electricity is often the classic example of the natural monopoly. Electricity cannot be stored and requires nearly instantaneous production and consumption after conducting across transmission wires. As such, a transmission network must be built and regulated to accommodate the delivery of electrical power to households and businesses. The power industry requires government regulation in order to fully realize the externalities of the network effects in building the power grid.
Historically, governments have allowed for the existence of local monopolies that are vertically integrated. In the power sector, this vertical integration takes the form of a single utility
company that owns and operates nearly all levels of electricity production and delivery – the utility owns the generation that produce electric power, the high-voltage transmission lines that deliver power to the distribution grid, which provides lower-voltage power to commercial and residential customers and whose billing and service is also managed by the same utility.
Utilities are required to submit their costs of operations and rationale for investments to
regulators each year through rate cases, an annual ritual of give-and-take and negotiation. In the United States these regulatory boards, often called public utility commissions (PUCs) sit at the state level. These regulatory boards decide on a case-by-case basis for each utility what customer usage charges to allow the distribution portion of the energy delivery system. In regulated regions, these rates, or tariffs, are often based on a “cost-plus” or “CPI – X” formula
which allow utilities to recover their costs and make a slight return while possibly providing for limited incentives. Other than the sale of power exports from their area, the primary source of revenue for utilities is through these tariffs.
Regulated utilities make investments based on anticipated future power needs and the ability to demonstrate to PUCs that building a generator serves the reliability and affordability of the utility’s customers. An entire suite of factors are considered – environmental impact, fuel diversity and security, and social equity – but the primary criterion is often cost. Traditional regulation has encouraged utilities to be risk-averse and conservative; regulated utilities have traditionally not been promoters or early adopters of new power generation technology. Under regulation, utilities can only receive a set rate of return, so risky projects are discouraged when safer, but more expensive, projects exist where costs are guaranteed to be recovered.
2.5.2. Restructuring and Competitive Markets
In the last three decades in certain markets, the power sector has undergone significant market reform away from the traditional regulated vertical monopoly. These reforms often blend some set of liberalization and deregulation measures to form wholly restructured markets. Traditional utilities in these regions are split between companies that sell power from their own generation and distribution companies that purchase power and deliver it to end-users. Transmission infrastructure is owned privately, but control and responsibility of operating the grid is handed over to regional transmission operators (RTOs) or independent system operators (ISOs).
Multiple rationales exist for restructuring the power sector. By liberalizing the power sector and allowing many different companies to enter the electricity business rather than a single
monopoly, private capital from more companies can be used for power investments. Private investment also allows private firms to take on the risk of doing business in the power sector rather than having captive ratepayers shoulder risk, as regulated tariffs and cost recovery required. The introduction of additional firms should also allow for competition in the power sector, putting pressure to drive down costs and to make the sector more efficient. Finally, restructuring the market also can entail the introduction of local pricing through locational
marginal prices (LMPs) that provide price signals for investors. Higher LMPs provide incentives for building new generation or transmission in a particular region.
Figure 2-1 Status of Electricity Restructuring by State, September 20095
In areas where an ISO handles regional dispatching of generation, the prices within the market are determined by calculating an optimization problem for the system area for the lowest overall cost considering transmission and reliability constraints. The LMP in a particular zone is
considered the market clearing price and all electricity generation is paid at that price, even if their marginal costs are lower.
Merchant operators, as generation-only companies are often called, base their investment decision on whether they can profitably sell the electricity through a bilateral contract to a distribution utility or on the market. Captive electricity consumers are not exposed to the risk of poor investment decisions or pricey generation options as they were in regulated markets. However, in the absence of a guaranteed return, firms will only make investments that will be profitable.
5
2.5.3. Electricity Costs
The cost of electricity to consumers and producers are determined by many different factors, from the retail to the wholesale level. The price of wholesale electricity, or the electricity being traded and delivered on the transmission grid, can be established through long-term bilateral contracts where a supplier guarantees a fixed price to a customer. Prices for wholesale electricity can also be established through a marginal cost dispatch calculation, discussed in further detail in Section 3.2.
A large portion of the electricity cost in a fossil fuel power plant derives from the cost of fuel like oil, natural gas or coal. Fuel costs are passed onto the cost of electricity according to how efficient the power plant is, often expressed in heat rate (the amount of energy input versus the electrical energy output). Other variable costs for a power plant include the operating and maintenance cost and, for power plants operating under a carbon emissions price, the cost of emitting CO2. Fixed costs also include some set operating and maintenance but the largest
portion of fixed costs is in construction and capital costs, which are carried for the book life of the power plant.
Delivering electrical power entails transmission costs which can involve payments to
transmission infrastructure owners. Power lost to resistance in the transmission lines is minimal – generally only a few percent – but the generation company is responsible for making up the shortfall. The distribution network contains even more resistance losses than the transmission system and requires more maintenance and administration to bill customers and provide service.
2.6. Area of Interest – Western Interconnection
The study area is the Western Interconnection of the United States. The Western Interconnection is one of three electrical interconnections in the United States, along with the Eastern
Interconnection and the Texas Interconnection. Interconnections are isolated electric networks that contain special transmission connections in order to ensure synchronization within each region.
Figure 2-2 Electrical Interconnections in the United States and Canada6
Electricity reliability and coordination is managed by the Western Electricity Coordinating Council (WECC). The area under their authority contains the western US states of Washington, Oregon, California, Nevada, Idaho, Utah, Colorado, Arizona, and parts of Montana, Wyoming, South Dakota, Nebraska, New Mexico, and Texas. The Canadian provinces of Alberta and British Columbia are also under WECC’s purview as well as northern Baja California, Mexico.
Other organizations and areas that the Western Interconnection contains include the West Coast Regional Carbon Sequestration Partnership (WESTCARB), one of the DOE’s regional
sequestration initiatives to advance the state of the art in CCS. The Western Interconnection is largely a regulated area with California ISO (CAISO) as the only deregulated market. The Bonneville Power Authority (BPA) controls a large portion of the hydroelectricity capacity in the northwest.
The Western Interconnection features important regional differences among its thirty-seven balancing authorities. The northwestern region is dominated by inexpensive hydropower, a resource highly dependant on seasonal weather for filling water reservoirs. The mountain states
6
along the Rockies utilize the cheap, abundant resources from Montana and Wyoming: coal. As such, in Arizona, New Mexico, Utah and Nevada many of the power plants are coal-fired and also relatively polluting in their emissions. In California, due to stricter environmental regulations and higher population densities, utilities have emphasized natural gas in order to fulfill demand.
3. METHODOLOGY
The methodology used in this thesis is outlined below and further explained in this chapter: 1) Choose a model for the simulation calculations. (Section 3.1: Dispatch Model)
2) Establish performance and economic metrics like capacity factors to be measured from dispatch simulations. (Section 3.2: Dispatch and Capacity Factor Calculation)
3) Estimate individual marginal costs for all existing power plants. (Section 3.3: Marginal Costs)
4) Determine attributes of individual hypothetical coal-fired CCS power plants. (Section 3.4: CCS Power Plants)
5) Decide locations for hypothetical CCS power plants to be inserted in the simulated electric power grid based on transmission, sequestration potential and other contextual factors. (Section 3.5: Discussion of Hypothetical Power Plant Sites)
6) Pick portfolio of sensitivity and market scenarios to use to explore CCS power plant dispatch. (Section 3.6: Simulation Scenarios Summary)
3.1. Dispatch Model
The model we have chosen to explore transmission constraints is an optimized power flow (OPF) model. The commercial software PowerWorld Simulator version 13 was selected with a
software add-on that also calculates OPF solutions. The data used to characterize the WECC study region will be described in detail below.
Several methods for calculating the dispatch of power system exist. The simplest is economic dispatch (ED). In ED, a system operator would line up generation by marginal cost, cheapest to most expensive, in merit order. The price on the electricity grid is set by the cost of the marginal unit of electricity. In order to serve load under ED, the system operator would use the set of cheapest power necessary to meet the demand. Given each generator’s capacity and marginal costs, ED is computationally simple. However, ED does not account for transmission constraints or redispatch due to congestion.
Optimized Power Flow (OPF) solves the power and circuit engineering problems to determine the level of current flow and voltages at all buses, the nodes of the electricity network, in the
power grid given transmission connections ,power injections and withdrawals. OPF will observe transmission line capacities so that lines will not be overloaded. The OPF algorithm will then look for the next cheapest available generation that will not violate transmission constraints and dispatch power plants “out-of-merit” order. The result is a dispatch that will serve load with the cheapest system cost given the constraints imposed by the transmission network.
Another more stringent type of dispatch is Security-Constrained Optimized Power Flow
(SCOPF). SCOPF uses OPF calculations to dispatch in the presence of transmission constraints but also is aware of security constraints. Security constraints concern the reliability of the electricity grid and contingency violations, such as an unexpected transmission line failure or generator outage. A SCOPF solution must also allow for these contingencies so that, for
instance, a failed transmission line will not overload other lines with rerouted power and cause a cascade that will take down the entire system. Security constraints thus ensure that n-1 or even
n-2 generators (situations with the sudden unplanned loss of one or two generators) will operate
reliably. We were unable to obtain definitions of security constraints in the WECC and thus relied on OPF calculations for capacity factor calculations.
Calculating a dispatch solution requires numerically solving a system of nonlinear equations. Given a system, an initial state and a set of constraints, there is no guarantee that a solution within tolerances can be found that does not violate a system of constraints. This is especially true given the large system that is used for our study area.
The WECC provided network data for the Western Interconnection. This data was a solved load flow case for the Western Interconnect on August 25, 2005, approximating an annual peak load. It represents the generation, transmission flow, and load for a moment during that date. The data include transmission connections, electricity generators and their capacities. The network data represent 58,000 miles of transmission, 190,000 MW of generating capacity, and 2,886
generators within the Western Interconnection. The power plants represented in the dispatch model network after cross-referencing are shown in Figure 3-1.
1000 MW coal natural gas nuke hydro wind other coal natural gas nuke hydro wind other 1000 MW 1000 MW
Figure 3-1 WECC Power Plants Represented in the Dispatch Model.
Color indicates power plant type (by fuel) and size corresponds to the size of the generator unit.
The network data can be considered a single solution in a moment of time, or a “snapshot” of the power grid at a particular instant. One consideration is that the load and generation distribution may be different for a hot summer afternoon than for an off-peak time during the shoulder months of spring and autumn. Also, updates to the transmission grid, such as additions in generation and transmission lines since 2005, will not have been modeled in the network data. However, the 2005 model data is a good approximation as the amount of additions in the past years have not been significant. (WECC Staff 2006)
In actual electricity operations, dispatch calculations like OPF are undertaken by the grid operator of a balancing authority. A regulated utility will act as their own balancing authority and will undertake their own economic dispatch and OPF calculation within their area to provide
load to the demand. In the Western United States, of the thirty-seven balancing authorities in the region, only one is a deregulated market which uses OPF to determine the power for a large group of distribution companies: CAISO. Our modeling effort is used to calculate the optimal generation for the entire Western Interconnection, in effect, pooling the power generation for the entire interconnection.
The assumption that the entire interconnection is one large dispatchable area is used for several reasons. First, the power from other balancing authorities outside of a hypothetical power plant’s is crucially important. Power is transmitted over long distances in the interconnection to reach load, possibly causing congestion for other generation. Some of the scenarios modeled, most notably the carbon pricing scenario, will drastically affect the dispatch of many power plants in other balancing authorities, requiring all power plants to be redispatched. Second, due to data limitations, it is unclear what power plants are under which balancing authorities’ control. Certain power plants within other areas may be owned and operated by other entities. Finally, an OPF dispatch calculation will calculate the most economical method to distribute power
generation, representing a floor for overall cost.
It is important to keep in mind what marginal cost dispatch represents. A marginal cost dispatch calculation uses only the variable costs to determine whether a power plant should operate. It does not take into account the capital costs of building a power plant, a factor that can be considerable for earlyCCS power plants. The dispatch calculation is equivalent to making a short-run economic decision on whether to operate. It does not represent shutdown or startup decisions, which take into account longer run calculations, and can be considered as the case where a power plant has already been built and a utility merely wants to determine whether it will be worthwhile to operate. For the long-term investment decision, the financing calculation is determined from the simulations calculating revenue value.
In actual operation, a transmission grid operator in a deregulated market, like CAISO, only receives bids from power plants. That is, the inputs into dispatch calculation are values given by market participants, but the grid operator has no way of knowing whether the bids actually represent the short-run marginal costs as economic theory indicates participants should bid, or
whether they represent poor decision-making or even gaming of the system. Because the model used here uses estimated costs, actual operation may still produce different results.
3.2. Dispatch and Capacity Factor Calculation
A capacity factor is a metric of how much a generator’s capacity is used. If a power plant is running for every hour of the year at maximum capacity, that power plant would have a capacity factor of 100%. A 50% capacity factor may represent a power plant running for half of the hours at maximum capacity or at all of the hours at half capacity.
To run a full OPF calculation for each of the 8,760 hours in a non-leap year is computationally intensive and could take several days to calculate on a desktop computer. In order to run
simulations for each individual site and for many different policy and price scenarios, a stratified sampling approach was applied in which the annual hourly load was sampled at percentiles of the load duration curve, as well as the annual minimum and maximum load, for a total of 102 samples.
A dispatch calculation was made at each of these system load levels, and capacity factors were calculated from these dispatch calculations. This procedure ensured that that capacity factors are taken from a sample that takes into account the daily, weekly and seasonal variations in the load that occurs in regular electricity usage. For each new carbon price, fuel price, CCS plant
location, and CCS plant characteristic, a capacity factor was calculated by running 102 samples for each of these scenarios. An example of a capacity factor calculation can be found in
Appendix A: Dispatch and Capacity Factor Calculation Example.
There is an important caveat in understanding what capacity factor calculations represent. A calculated 100% capacity factor would not realistically result in a power plant running at full capacity for the entire year as there since any power plant is not fully 100% available and requires outages for critical maintenance. A 100% power plant capacity factor then would represent that the power plant could profitably run for all hours of the year at maximum capacity.
